Biogas upgrading through kinetic separation of carbon dioxide and methane over Rb- and Cs-ZK-5 zeolites

Transcription

1 Electronic Supplementary Material (ESI) for RSC Advances. This journal is The Royal Society of Chemistry 2014 Supporting Information for: Biogas upgrading through kinetic separation of carbon dioxide and methane over Rb- and Cs-ZK-5 zeolites Tom Remy, a Elena Gobechiya, b David Danaci, c Sunil A. Peter, a Penny Xiao, c Leen Van Tendeloo, b Sarah Couck, a Jin Shang, c Christine E.A. Kirschhock, b Ranjeet K. Singh, c Johan A. Martens, b Gino V. Baron, a Paul A. Webley, c and Joeri F.M. Denayer a* a Department of Chemical Engineering, Vrije Universiteit Brussel, Pleinlaan 2, 1050 Brussels, Belgium b Centre for Surface Chemistry and Catalysis, KU Leuven, Kasteelpark Arenberg 23, 3001 Leuven, Belgium c Department of Chemical and Biomolecular Engineering, The University of Melbourne, 3010 Victoria, Australia * Corresponding Author: Joeri F. M. Denayer joeri.denayer@vub.ac.be Tel.: Fax:

2 Table of contents S1 Detailed synthesis procedure of K-ZK S2 Parameter calculation/estimation for breakthrough simulations... 4 S3 PSA Simulations... 7 S4 XRD patterns S5 SEM results S6 Rietveld refinements S7 Dual-site Langmuir fits S8 Breakthrough data S9 Notation S10 References S2

3 S1 Detailed synthesis procedure of K-ZK-5 The synthesis was based on the procedure as disclosed by Verduijn et al. 1 The synthesis system is SiO 2 - Al 2 O 3 - KOH - Sr(NO 3 ) 2 - H 2 O with the following molar composition 5 SiO 2 - Al 2 O KOH Sr(NO 3 ) 2-78 H 2 O. First, 4.5g H 2 O (Milli-Q) are added to a polypropylene bottle (Nalgene, 100 ml) with 2.684g KOH (VWR, 85.9%) and 1.62g Al(OH) 3 (residue at 1373 K: 62-67%, BDH Chemicals Ltd.). This unclosed bottle is heated, while stirring, for about 20 min in an oil bath at 378 K until complete dissolution of Al(OH) 3. Afterwards, the bottle is weighed in order to determine the mass loss due to water evaporation during the heating. In a second polypropylene bottle (Nalgene, 100 ml), g H 2 O (Milli-Q) plus the amount lost of water during the heating is added to g Ludox HS-40 (Sigma-Aldrich, 40 wt% suspension in H 2 O) under stirring. A third polypropylene bottle (Nalgene, 100 ml) is used to mix 4.5 ml H 2 O (Milli-Q) with 0.050g Sr(NO 3 ) 2 (VWR, 98 %) under stirring. Afterwards, the content of the third polypropylene bottle is added drop by drop to the second one. Subsequently the mixture in the second bottle is added drop by drop to the first polypropylene bottle. The resulting gel is transferred to a Teflon liner, put in an autoclave, and kept in a hot air oven at 423 K for at least 96 h. Finally the sample was vacuum-filtered and washed several times with deionized water and dried overnight at 353 K. With this procedure about g K-ZK-5 can be obtained per Teflon liner. S3

4 S2 Parameter calculation/estimation for breakthrough simulations Conservation equations The axial dispersion coefficients for CO 2 and CH 4 were calculated from the Wakao et al. correlation: 2 D ax,i = D m,i ( E ε Re Sc) B The Reynolds (Re) and Schmidt (Sc) numbers for CO 2 and CH 4 were calculated as follows: Re = ρ g v d p µ g in Sc = µ g ρ g D m in E 0 represents the stagnant contribution to the axial dispersion and a value of 0.23 has been recommended. 3 The temperature dependence of the molecular diffusion coefficient is given by: 3 T D m,i T = D m,i Tref T ref 1.75 Molecular diffusion coefficients at a reference temperature of 293 K were taken from Delgado et al. 3 The obtained axial dispersion coefficients for CO 2 and CH 4 are 7.1 x 10-5 m 2 /s and 7.7x 10-5 m 2 /s respectively. The pellet density is calculated from the following equation 4 : ρ p = (1 ε p ) ρ c The crystal density ρ c can be estimated from framework density (given on the IZA website) and mass of unit cell, whose composition has been determined by ICP. For the simulations, the following typical values for ε p and ε B were used: S4

5 ε p = 0.5 ε B = 0.4 This results in a typical value of 0.7 for the total porosity, where the total porosity was calculated via 4 : ε t = ε B + (1 ε B ) ε p Momentum balance - Ergun equation The density of the gas mixture is obtained from: 5 ρ g = j y j MW j P R g T The viscosity of the gas mixture is obtained from the Wilke equation: 6 µ g = µ mix = i j µ i y i y j Φ ij in which: 1+ µ MW i j µ i MW i Φ ij = µ mix = 8 1+ MW i MW j 1/ 4 2 This function is dimensionless and equal to 1 when i = j. The pure component viscosities are calculated from the Chapmann-Enskog equation: 6 µ g = MW i T σ c 2 Ω µ Pure component values for σ c and Ω µ can be found in book of Bird et al 6. S5

6 Boundary conditions The boundary conditions consist of the well-known Danckwertz boundary conditions 4 : [ v C i ] in = [ v C i ] z=0 D ax,i C i dz z=0 C i dz z=l = 0 In addition, the exit pressure of the bed is fixed at 1 bar or any pressure given as setpoint to the backpressure regulator. S6

7 S3 PSA Simulations For comparative purposes a simple 6-step PSA process with 2 beds for separation of an equimolar CO 2 /CH 4 mixture was used (Figure S1). The separation consists of the following steps: feed (bed 1) via vacuum evacuation (bed 2) (step 1)- pressure equalization via the product ends between bed 1 and bed 2 (step 2) - depressurization (bed 1) via repressurization with feed (bed 2) (steps 3 + 4) - pressure equalization via the product ends between bed 1 and bed 2 (step 5) - repressurization with feed (bed 1) via depressurization of bed 2 (step 6). 1! 2! 3! 4! 5! 6! 1! 2! 1! 2! 1! 2! 1! 2! 1! 2! 1! 2! Figure S1. Six-step PSA process for separation of an equimolar CO 2 /CH 4 mixture. The function of each step for bed 1 in the context of CO 2 /CH 4 separation is briefly explained below: 4, 7 1. Feed with raffinate withdrawal Withdrawal of raffinate at high purity (CH 4 production step). In a PSA process, the raffinate is by definition the weakest adsorbed component (CH 4 in this case). The feed gas containing 40% CO 2 balanced with CH 4 enters the system at high pressure. CO 2 is adsorbed and highly pure CH 4 (CH4 product) exits from the bed. S7

8 2. Pressure equalization After the second bed completed the evacuation step and the first the feed step, both beds are connected via their product ends until the pressures in both beds get equal. Since the second bed is now partly pressurized with gas from the first bed, energy is being conserved. At the same time, separative work is conserved as the stream from bed 1 to bed 2 is depleted in the strongly adsorbed component (CO 2 ). This step will benefit a high CO 2 concentration product in the next steps. 3. Depressurization Achieving high purity of both extract and raffinate products. This is the preferred regeneration option when the extract (CO 2 ) is strongly adsorbed. In a PSA process, the extract is by definition the strongest adsorbed component (CO 2 in this case). As the pressure in the bed is released, the adsorbed CO 2 adsorbed is released into the gas phase at a low pressure. 4. Evacuation (vacuum regeneration) Same as for step 3, with concurrent raffinate withdrawal from bed 2. Because CO 2 is strongly adsorbed, a very low pressure is required to remove CO 2 from the bed using a vacuum pump. 5. Pressure equalization See step 2, but now beds 1 and 2 are interchanged. 6. Repressurization Enrichment of the less adsorbed species (CH 4 ) in the gas phase at the product end (production of CH 4 with high purity). The bed is repressurized with feed gas until the pressure reaches that in step 1. As a result, pure CH 4 (raffinate) and high concentration CO 2 (extract) can be generated continuously with the cyclic process. The parameters for the reference PSA process are given in Table S1 - S2. S8

9 Table S1. Parameters for PSA separation of CO 2 /CH 4 mixtures. Parameter Value ε B (-) 0.37 ε p (-) 0.6 ρ p (kg/m 3 ) 641 d p (m) C p (J/(kg.K)) 1000 Wall thickness (m) 3.5 x 10-3 C pw (J/(kg.K)) 477 ρ W (kg/m 3 ) 7900 Bed length (m) 1.0 Bed internal diameter (m) 1 x 10-2 T ambient (K) P ambient (bar) T initial (K) P initial (bar) 2.0 y CO2initial 0.5 y CH4initial 0.5 For comparative purposes, the same value for the bed porosity, pellet porosity, adsorbent density, pellet diameter and adsorbent heat capacity were assigned to each adsorbent. Mixture properties were calculated in the same way as for the breakthrough simulations (see section S1). S9

10 Table S2. Step times and pressures at the end of each step for PSA separation of an equimolar CO 2 /CH 4 mixture on different adsorbents (PE = pressure equalization, depres. = depressurization, ev. = evacuation, repres. = repressurization). 13X F in = mol/s (2 Nl/min) t feed = 55s t PE1 = 5s t depres. = 50s t ev. = 55s t PE2 = 5s t repres. = 50s P feed = 2 bar P PE1 = 0.92 bar P depres. = 0.15 bar P ev. = 0.10 bar P PE2 = 0.9 bar P repres. = 2.02 bar Rb-ZK-5 F in = 5.0 x 10-4 mol/s (0.67 Nl/min) t feed = 30s t PE1 = 5s t depres. = 100s t ev. = 30s t PE2 = 5s t repres. = 100s P feed = 2 bar P PE1 = 1.05 bar P depres. = 0.1 bar P ev. = 0.1 bar P PE2 = 1.05 bar P repres. = 2.02 bar Cs-ZK-5 F in = 2.0 x 10-4 mol/s (0.67 Nl/min) t feed = 25s t PE1 = 5s t depres. = 100s t ev. = 25s t PE2 = 5s t repres. = 100s P feed = 2 bar P PE1 = 1.0 bar P depres. = 0.05 bar P ev. = 0.05 bar P PE2 = 1.0 bar P repres. = 2.02 bar S10

11 S4 XRD patterns Figure S2. XRD pattern of Rb-ZK-5. Figure S3. XRD pattern of Cs-ZK-5. S11

12 S5 SEM results Figure S4. SEM image from as-synthesized K-ZK-5. S12

13 S6 Rietveld refinements a) b) Figure S5. Rietveld plots for the XRD profile of Cs-ZK-5; R wp = 5.51, R p = 4.21%, χ 2 = 1.40; observed (black dots) and calculated (red solid line) patterns, the difference plot (blue bottom curve) and positions for Bragg-reflections (purple vertical bars - Cs-ZK-5; light blue vertical bars - chabazite): a) low-angle region; R F 2 = 4.43%; b) high-angle region; R F 2 = 14.80%. a) b) Figure S6. Rietveld plots for the XRD profile of Rb-ZK-5; R wp = 5.99, R p = 4.72%, χ 2 = 1.59; observed (black dots) and calculated (red solid line) patterns, the difference plot (blue bottom curve) and positions for Bragg-reflections (purple vertical bars - Rb-ZK-5; light blue vertical bars - chabazite): a) low-angle region; R F 2 = 3.90%; b) high-angle region; R F 2 = 12.17%. S13

14 Figure S7. Location of the M1-site within Rb-ZK-5. S14

15 S7 Dual-site Langmuir fits Table S3. Dual-site Langmuir constants for CO 2 and CH 4 on Rb-ZK-5 and Cs-ZK-5 Rb-ZK-5 CO 2 CH 4 q 1sat (mol/kg) 2.08 ± ± 0.04 q 2sat (mol/kg) 2.28 ± ± 16 b 10 (1/kPa) (3.0 ± 2.9) x (6.3 ± 3.1) x 10-5 Q 1 (J/mol) (46.4 ± 2.5) x 10 3 (11.1 ± 1.2) x 10 3 b 20 (1/kPa) (7.19 ± 0.14) x (3.1 ± 1.5) x Q 2 (J/mol) (54.7 ± 5.1) x 10 3 (102.0 ± 0.3) x 10 3 b 298 K (1/kPa) 4.0 x x 10-3 b 298 K (1/kPa) x 10-6 Cs-ZK-5 CO 2 CH 4 q 1sat (mol/kg) 2.55 ± ± 6.2 q 2sat (mol/kg) 2.16 ± ± 0.22 b 10 (1/kPa) (6.4 ± 3.7) x 10-7 (8.56 ± 0.18) x 10-8 Q 1 (J/mol) (18.8 ± 1.6) x 10 3 (18.2 ± 2.5) x 10 3 b 20 (1/kPa) (1.9 ± 2.3) x 10-8 (1.32 ± 0.92) x 10-6 Q 2 (J/mol) (38.9 ± 3.3) x 10 3 (21.2 ± 2.0) x 10 3 b 298 K (1/kPa) 1.3 x x 10-4 b 298 K (1/kPa) x 10-3 S15

16 S8 Breakthrough data 1.0! 0.8! F/F 0! 0.6! 0.4! 0.2! 0.0! 0! 1! 2! 3! 4! Time (min)! Figure S8. Pure CH 4 breakthrough profile on Rb-ZK-5 at 308 K and 1 bar (F TOT = 10 NmL/min). 1.0! 0.8! F/F 0! 0.6! 0.4! 0.2! 0.0! 0! 1! 2! 3! 4! Time (min)! Figure S9. Pure CH 4 breakthrough profile on Cs-ZK-5 at 308 K and 1 bar (F TOT = 10 NmL/min). S16

17 S9 Notation b = Henry constant (kpa -1 ) b 0 = Henry pre-exponential factor (kpa -1 ) C = concentration (mol/m 3 ) C p = pellet heat capacity (J/kg/K) C pw = wall heat capacity (J/kg/K) D ax = axial dispersion coefficient (m 2 /s) D m = molecular diffusion coefficient (m 2 /s) d p = particle diameter (m) E 0 = stagnant contribution to axial disperison (-) F = flow rate (Nl/min or Nml/min) L = column length (m) MW = molecular weight (kg/kmol) Re = Reynolds number (-) P = pressure (Pa) R g = gas constant (J/mol/K) Q = Langmuir adsorption heat (J/mol) q sat = saturation loading (mol/kg) Sc= Schmidt number (-) T = temperature (K) v = interficial velocity (m/s) y = mole fraction (-) z = axial position (m) ε B = bed porosity (-) ε p = pellet porosity (-) ε t = total porosity (-) S17

18 σ c = collision diameter (Å) Ωµ = collision integral ρ c = crystal density (kg/m 3 ) ρ g = gas density (kg/m 3 ) ρ p = pellet density (kg/m 3 ) ρ W = wall density (kg/m 3 ) µ g = gas viscosity (Pa.s) S18

19 S10 References 1. U.S. Pat., 4,994,249, N. Wakao, S. Kaguei and H. Nagai, Chem. Eng. Sci., 1978, 33, J. A. Delgado, M. A. Uguina, J. L. Sotelo, B. Ruiz and J. M. Gomez, Adsorption, 2006, 12, D. M. Ruthven, S. Farooq and K. S. Knaebel, in Pressure Swing Adsorption, VCH Publishers, Inc., New York, R. S. Todd, G. B. Ferraris, D. Manca and P. A. Webley, Comput. Chem. Eng., 2003, 27, R. B. Bird, W. E. Stewart and E. N. Lightfoot, in Transport Phenomena, John Wiley & Sons, Inc., New York, C. A. Grande and A. E. Rodrigues, Ind. Eng. Chem. Res., 2007, 46, S19